Most of the plotting methods in Bokeh API are able to receive data source parameters through ColumnDatasource object. It makes sharing data between plots and ‘DataTables’.

A ColumnDatasource can be considered as a mapping between column name and list of data. A Python dict object with one or more string keys and lists or numpy arrays as values is passed to ColumnDataSource constructor.

Example

Below is the example

from bokeh.models import ColumnDataSource
data = {''x'':[1, 4, 3, 2, 5],
   ''y'':[6, 5, 2, 4, 7]}
cds = ColumnDataSource(data = data)

This object is then used as value of source property in a glyph method. Following code generates a scatter plot using ColumnDataSource.

from bokeh.plotting import figure, output_file, show
from bokeh.models import ColumnDataSource
data = {''x'':[1, 4, 3, 2, 5],
   ''y'':[6, 5, 2, 4, 7]}
cds = ColumnDataSource(data = data)
fig = figure()
fig.scatter(x = ''x'', y = ''y'',source = cds, marker = "circle", size = 20, fill_color = "grey")
show(fig)

Output

Instead of assigning a Python dictionary to ColumnDataSource, we can use a Pandas DataFrame for it.

Let us use ‘test.csv’ (used earlier in this section) to obtain a DataFrame and use it for getting ColumnDataSource and rendering line plot.

from bokeh.plotting import figure, output_file, show
import pandas as pd
from bokeh.models import ColumnDataSource
df = pd.read_csv(''test.csv'')
cds = ColumnDataSource(df)
fig = figure(y_axis_type = ''log'')
fig.line(x = ''x'', y = ''pow'',source = cds, line_color = "grey")
show(fig)

Output